Use of color in the office and business environment is growing rapidly. In the 90s, color is becoming easier and less expensive to use and its application is moving from isolated color pages toward integrated color documents. Moreover, in parallel with this trend, color in documents is driving a paradigm shift from a text orientation to a pictorial and graphic one. Business applications cover a range of quality requirements for both full and highlight color documents. Non-impact printers for these applications are being manufactured based on numerous technologies, including xerography, thermal transfer, ink jet, thermal dye diffusion and even photography. We compare and contrast these technologies in terms of characteristics such as registration, image structure, color gamut, and stability. We then consider the systems impact of customer requirements such as ease of use and consistency of results across printers, demonstrating a need to hide the printer technologies and their differences from the system users. A need for standards in both the description of color documents in device independent form and in other aspects such as viewing conditions is discussed in this context.

Hardcopy is a picture produced from electronic signals involving visual information, although its scope has been gradually expanding to cover the whole permanent visual images printed on the recording media. The coloration of hardcopy started in the middle of the 1960s. During the past three decades, various kinds of hardcopy systems have developed. The output pictures of those systems have extended from documental pictures expressing simple textural images of numerals and words for business use (documental hardcopy) to aesthetical images of impressive color photograph-like prints (pictorial hardcopy). The main requirements of the documental hardcopy systems are to realize the dry and fast processing, safe and maintenance-free performances on the equipment. Those of the pictorial hardcopy systems are to realize the high fidelity or favorable tone and color reproductions on the output image. At the first stage, the images on every hardcopy had shown the documental contents. The development of the pictorial hardcopy started at the beginning of the 1980s. Recently, the advanced technologies for the pictorial hardcopy have fed back into the documental hardcopy system and various types of color equipment have appeared on the market. This paper deals with the current competition between the pictorial and the documental color hardcopy systems.

The properties of conventional, ordered dot-pattern generation techniques for bi-level halftone representation are examined and compared with the properties of error-diffusion based, disordered dotpattern generation algorithms. The various processing steps necessary for the adaptation of the disordered halftone pattern-generation technique to digital image hardcopy reproduction with non-ideal computer- output printing devices are described. It includes procedures for spatial distribution of thresholding errors, suppression of dot-density artifacts and compensation for dot overlap. These procedures represent the core of the Multiple-Error Correction Computation Algorithm (MECCA), the objective of which is to linearize the non-ideal printing process to minimize the loss or shift of tonal gradations

Single-pass highlight color printers (black plus one color) are advantaged over multi-pass systems in that image throughput is not sacrificed in obtaining the additional color. Furthermore, single exposure systems are advantageous since perfect registration is inherently assured. In this paper we will review several approaches to single-pass highlight color systems and then focus on the recently announced Xerox highlight color printer.

Over much of the last half century, electronic technologies have played an increasing role in the prepress production of film and plates prepared for printing presses. The last decade has seen an explosion of technologies capable of supplementing this production. The most outstanding technology infusing this growth has been the microcomputer, but other component technologies have also diversified the capacity for high-quality scanning of photographs. In addition, some fundamental software and affordable laser recorder technologies have provided new approaches to the merging of typographic and halftoned photographic data onto film. The next decade will evolve the methods and the technologies to achieve superior text and image communication on mass distribution media used in the printed page or instead of the printed page. This paper focuses on three domains of electronic prepress classified as the input, transformation and output phases of the production process. The evolution of the component technologies in each of these three phases is described. The unique attributes in each are defined and then follows a discussion of the pertinent technologies which overlap all three domains. Unique to input is sensor technology and analogue to digital conversion. Unique to the transformation phase is the display on monitor for soft proofing and interactive processing. The display requires special technologies for digital frame storage and high-speed, gamma-compensated, digital to analogue conversion. Unique to output is the need for halftoning and binary recording device linearization or calibration. Specialized direct digital color technologies now allow color quality proofing without the need for writing intermediate separation films, but ultimately these technologies will be supplanted by direct printing technologies. First. dry film processing, then direct plate writing, and finally direct application of ink or toner onto paper at the 20-30 thousand impressions per hour now achieved by offset printing. In summary, a review of technological evolution guides industry methodologies that will define a transformation of workflow in Graphic Arts during the next decade. Prepress production will integrate component technologies with microcomputers in order to optimize the production cycle from graphic design to printed piece. These changes will drastically alter the business structures and tools used to put type and photographs on paper in the volumes expected from printing presses.

Although electrophotography has been used extensively for years in copying, desktop and department printing, and certain in-plant printing operations, this non-impact printing technology has only recently found applications in high-volume commercial printing of products such as books and documentation. Several factors have influenced the growth of electrophotography in commercial printing: (1) the gradual reduction in electrophotographic printing costs achieved through increased equipment reliability, higher productivity, and economies of scale and learning, (2) the increasing availability of electronic pages in standard, device-independent formats such as PostScript, (3) increasing standardization and connectivity of electronic printing systems, and (4) steady increases in electrophotographic print quality. As a result of these technology advances, electrophotography has begun to complement offset as a demand printing option in certain markets and product categories. The successful applications of electrophotography in commercial printing exploit the value of demand printing along three primary dimensions: (1) time management, (2) customization, and (3) inventory management. This paper will illustrate these value dimensions, using several examples of products produced electronically. Finally, the author will review the critical factors pacing the evolution of electrophotographic demand printing in commercial printing over the next few years.

One exposure, trilevel xerography provides a novel means to produce, single pass, highlight color prints. In this scheme, a ROS creates an imagewise three level, unipolar latent image in which the color information is encoded in the discharge level. The highest charge regions represent one color, and the lowest charge regions a second color. An intermediate charge level, approximately midway between the high and low charge, serves as the background reference and is not printed. The components of the composite two color latent image are developed sequentially using charged area and discharged area development and bipolar developers. Electrostatic transfer is enabled by converting the developed bipolar image to a unipolar image using a pretransfer charging step. Because of the nature of the latent image, a trilevel development system must satisfy demands substantially beyond those required in conventional xerography. This paper will discuss these requirements.

Single pass highlight color printing with tri-level xerography has some unique capabilities, but also has some inherent difficulties which needed to be overcome before launching the new Xerox 4850 Highlight Color Laser Printing System. These capabilities and potential pitfalls are discussed below, with the aid of a computer model of tri-level xerography.

Electrographic printing with a latent image generation by direct dielectric charging offers the advantage of continuous toning. Such continuous toning can be used for both full grey scale and full colour printing. For high quality imaging, in both cases, the evenness of charge distribution and image edge sharpness and smoothness become requirements. The latent image formed by a matrix of electron beams is degraded by charge-to-charge interactions. Outcomes of such interactions are the blooming effect, the image deflection, and compression which together cause non-uniformity in the horizontal charge distribution. Variability in the horizontal charge distribution is periodic and produce the vertical banding with DPI/2 frequency. To suppress the banding defect, printing with variable dot order is suggested. The charge distribution uniformity can also suffer from the variability of the charge emission from individual charge generating sites caused by the printhead manufacturing tolerances. Liberation of the print quality from the printhead structural imperfections seems to be feasible using the latent image potential rather then the electron beam timing for the deposited charge density limitation. The principal structural arrangement of such a printing system is discussed.

Ion flow printing technologies, one of which is ionography, have possibilities of pixel tone control. The conditions of pixel tone control have been studied, and high quality color printing has been achieved. Pixel tone is controlled by the time duration of ion projection. The relation between the time duration and the amount of deposited charge show quite good linearity. Electrostatic latent images are developed by liquid developer and color prints have been obtained by developing cyan, magenta and yellow images. The application of the ion flow printing technology to high quality electronic imaging systems, color photograph image filing system and color facsimile, is discussed.

A new statistical method for analyzing corotron charging in a xerographic process has been investigated. Four parameters of corotron charging and four settings of each parameter are selected to analyze the effect of corotron charging. In the conventional method, if one parameter and one setting are selected in one run, it needs 256 runs to complete the total experiment. If the L16 table of Taguchi method is used, it only needs 16 turns to complete the run test.

Recent technology progress in organic photoconductor OPC is remarkable. Various materials have been developed and now a variety of OPC are available in the market. Some of them are very high sensitive in visible region. High sensitive OPC in near IR wavelength have been developed for the laser beam printer application. They have also enough durability. Now we can select a suitable OPC for most of the applications of the electrophotography. Historical review and recent technological trends are shown in this paper.

Recent needs ofhigher copy quality with dry xerography led to aggressive pursuit of smaller particle size toners also in the toner technology areas. This is based on a beliefthat copy quality will be improved as the toner particle size is reduced. On the other hand, the toner size reduction has been known to cause difficulties in the xerographic process, such as poor transfer properties, for example. The development and transfer steps in the xerographic system depend on detachment of charged toner particles by the electric field which has to overcome the adhesion force to carrier beads and photoreceptor. In order to find out small toner well matching the xerographic process, understanding of adhesion forces of charged toner particles is necessary. We have evaluated the detachment field of charged toner particles varying the size, composition materials and surface structure. The results showed that the detachment field was dominated by the electrostatic force which in turn depended on the toner surface roughness. This roughness can be related to the actual contact area with the substrate surface. 7 ?m toners designed according to the above findings gave good transfer properties.

Magnetic toners in a hybrid-component developing system and full-color toners in a twocomponent developing system have been investigated. Magnetic toners in different steps of Xerographic processes were selected to measure the tribo-charge and the particle size distribution. The average particle size of the residual toners is larger than that of the original toners, and the toner with small particle size and large q/d value is eaiser to be developed. For color toners, the mass per unit area are in the range of 1 ?g/cm2~ 3.5 ?g/cm2 and the value of charge-to-mass ratio are in the range of 2 ?c/g~ 8.1 ?c/g.

We have evaluated some azo pigment carrier generation materials in a layered organic photoreceptor in terms of electrophotographic sensitivity and color reproduction, and found a dihydro-thiadiazepine derivative azo pigment which shows high sensitivity and excellent color response. Fixing carrier generation material(CGM) at this dihydrothiadiazepine derivative and selecting some conventional carrier transport materials(CTM), we have investigated effects of carrier mobility in a carrier transport layer(CTL), carrier injection efficiency and carrier generation in CGL on the electrophotographic sensitivity. If we assume carrier generation in carrier genaration layer(CGL) is well represented by carrier peak current, the electrophotographic sensitivity has been found to show the closest corelation with the carrier peak current. The carrier peak current also shows close corelation with carrier life time in CGL which is strongly influenced by CTM. We propose a mechanism that explains how CTM affects electrophotographic sensitivity of layered organic photoreceptor.

This paper discusses the use of Rietveld analysis to solve crystal structures of titanyl phthalocyanines. Rietveld analysis is based on whole pattern fitting in which every point in the x-ray powder diffraction pattern is used as a measure of one or more Bragg diffraction peaks. Thus the refinement of relatively complicated crystal structures from x-ray patterns containing a relatively small number of resolved diffraction peaks is enabled. Various profile-fitting functions used in Rietveld analysis were parameterized and tested against known crystal structures of type I and type II titanyl phthalocyanine. It was found that a split Pearson VII function was found to best correct for preferred orientation effects observed in the x-ray patterns. The final goodness-of-fit parameters were R(Bragg) = 0.17 and 0.13 for type I and type II structures, respectively. A computer program was used to generate several unit cells for type IV titanyl phthalocyanine. These unit cells were tested by stereochemical packing analysis to first determine which unit cells allowed for good intermolecular packing arrangements. Energy minimized models were then used as phasing models for Rietveld refinement. A triclinic structure with space group P-1 having an x-ray goodness-of-fit parameter R(Bragg) = 0.24 was proposed as the most probable crystal structure for type IV titanyl phthalocyanine. The unitcell parametersarea = 1.083 nm, b = 1.312 nm, c = 0.996 nm, alpha = 72.28 degrees, beta = 77.25 degrees and gamma = 104.48 degrees. There are two molecules in the unit cell related by a center of inversion.

A color copy may exhibit more severe curl than a black-and-white copy due to profound shrinkage difference between multiple toner layers and the base paper after the fusing process. An untreated curled copy may cause paper handling problems and it is unacceptable to customers. This paper analyzes factors contributing to copy curl developed after the fusing process with an emphasis on the stiffenss and shrinkage interactions between the toner layer and the paper. Specifically, a bi—material curl model of toner— paper composite is formulated in terms of ratios of toner/paper thickness, modulus of elasticity and expansion coefficient. The curl model regards the temperature and moisture changes as the driving forces for the curling of a fused copy. Besides predicting away—from—image and toward—image curls, the model can be used to determine requirements of toner modulus,layer thickness and fusing temperature to meet specified curl limits of fused copies.

The conventional thermal paper utilizes an electron-donating colorless dye and an electron-accepting developer, which react to give colored image. To improve the image stability, various kind of methods have been proposed. However, there has been no study on the prediction of the image stability. We found that under some defined condition of thermal acceleration test,Arrhenius Plot of the fading gives a straight line, and that the predicted value gives good coincidence with the value obtained by actual aging. This report presents the outline of the method for prediction and the coincidence of the result of the test method with that obtained by actual aging.

The first public demonstration of thermal ink jet printing was done by Canon in 1981 and the first thermal ink jet product, ThinkJet, was introduced by the Hewlett-Packard Company in 1984. Since then, this powerful printing technology has assumed a strong presence in the market. In this discussion, we will first briefly review the printer market, the increasing role thermal ink jet is playing in this arena, as well as the reasons for its success.

The velocity and diameter of ink drops are important parameters in ink-jet printing technology. They affect print quality and the control of the gray scale. We report here an exploratory experiment in which the velocity and diameter of drops from thermal ink-jet print heads were measured nonintrusively, simultaneously and in real time. A laser phase Doppler particle analyzer was used in the experiment. The velocity was deduced from the Doppler signal and the diameter from the spatial phase difference of the scattered intensity. From an analysis using geometrical optics, it was determined that the optimum condition for sizing absorbing spheres such as ink drops was that the signal detected be limited to externally reflected rays. The intensity of rays refracted as a fraction of rays reflected by ink drops could vary significantly over their size range and compounded size interpretation. By considering the scattered intensity, the analysis showed that the best placement of the detectors for the present application was at 90° scattering angle, using a perpendicularly polarized light source. In this configuration, twice refracted rays (no internal reflection) were absent. The detected signal was dominated by externally reflected rays. In order to measure ink drop diameter accurately, the complex refractive index m of several ink-jet inks was measured. A typical value was m = 1.37 — i 0.0049 at 0.6328?m wavelength. Monodisperse droplets produced with a piezo-electric squeeze tube drop generator were used to calibrate the instrument. With the optimized configuration the drop velocity and diameter of two commercial thermal ink-jet heads and an experimental prototype were measured.

The fundamental physico-chemical properties of hot melt inks are investigated. Potential problems exist in a number of key areas if efficient operation and print quality goals are to be achieved. These problem areas are discussed and practical solutions suggested.

Determining the color repeatability of images from color output devices is paramount for color-critical applications. Color should be described in coloriinetric values which are then used to make a probability statement about printed color repeatability. Experiments are described that determine color repeatability for a four color continuous ink jet printer using colorimetric methods and probability ellipses. CIELab measurements were taken from several ink jet prints which contained a standard test image. The test image consisted of an array of color patches with different colorant amounts. Two-dimensional ellipses, representing the 95% a*,b* confidence interval, were then computed for each color patch to establish color repeatability. Typical results showed that ellipse axes were not parallel to the a*,b* axes, suggesting that simple a*,b* tolerances would lead to an inaccurate repeatability statement.

Thermal Ink Jet (TIJ) depends on stable, consistent nucleation to create drops of uniform mass. The shape of the nucleating pulse is a contributing factor to the size and stability of the ejected drop. The size of the ejected drop can also be increased by the presence of a warming pulse immediately preceding the nucleating pulse. This paper summarizes the results of experiments designed to investigate these two phenomena.

Ensuring good quality colour reproductions requires regular calibration of all components in the system. However, this is not sufficient without proper characterisation and control of the printing press. It is a fundamental tenet of pre-press calibration that colour transforms be defined by the press characterisation and sufficient control be implemented to maintain consistent output using these transformations. Calibration may, therefore, be thought of as a 6 stage process: a) Stabilise and control the printing process. (This is increasingly becoming "standardised" via specifications such as SWOP) b) Characterise the press by printing and measuring suitable test images C) Match the proofing system to the press (for both soft and hard copy proofs) d) Stabilise (linearise) input scanners and output recorders e) Define the transformation required to produce high quality reproduct ions f) Control scanners and recorders to maintain the transform Procedures are discussed which may be used to achieve this calibration process and tools which have recently been developed to assist in this are described.

The conventional approach to designing an imagesetter is to initially specify the subsystem design parameters based on experience with similar existing systems. A breadboard model is then built and tested by extensively inspecting photocopy output samples. Following the results of these image quality performance tests, the system specification parameters are fine-tuned and a new breadboard model has to be built. Sometimes an additional breadboard iteration is required for tolerancing. We have developed an image fidelity simulation software package which, given the design parameters, simulates the photocopy output to a high degree of agreement with the output of actual imagesetters. Consequently we adopted a new approach to imagesetter system design in which we investigate the design parameter trade-offs and tolerances before the first breadboard model is built, thereby significantly reducing development time and cost. In this paper we define and discuss the image fidelity parameters, as would be measured on photocopy output. We further describe the image simulation software functionality and analyze the dependence of the image fidelity parameters on each of the system design parameters. We present a comparison between measured and simulated results and discuss the design and tolerancing procedure using the simulation approach.

The Generic Imaging Matching System (GIMS) provides an optimal systematic solution to any problem of color image processing in printing and publishing that can be classified as or modeled to the generic image matching problem defined. Typical GIMS systems/processes include color matching from different output devices, color conversion, color correction, device calibration, colorimetric scanner, colorimetric printer, colorimetric color reproduction and image interpolation from scattered data GIMS makes color matching easy for the user and maximizes operational flexibility allowing the user to obtain the degree of match wanted while providing the capability to achieve the best balance with respect to the human perception of color, color fidelity, and preservation of image information and color contrast. Instead of controlling coefficients in a transformation formula, GIMS controls the mapping directly in a standard deviceindependent color space, so that color can be matched, conceptually, to the highest possible accuracy. An optimization algorithm called Modified Vector Shading (MVS) was developed to minimize the matching error and to perform a "nearneighborhood" gamut compression. An automatic error correction algorithm with a multidirectional searching procedure using correlated re-initialization was developed to avoid "local minimum" failures. Once the mapping for color matching is generated, it can be utilized by a multidimensional linear interpolator with a small Look-Up-Table (LUT) implemented by either software, a hardware interpolator or a Digital-Signal-Processor (DSP)

A single spatial filter function closely replicating the contrast sensitivity of the visual system is used to predict the visual detection thresholds for narrow lines and dots and small amounts of edge blur. Both the predictions for lines and for edge blur are consistent with published data. Maximum differences with delta function lines and dots yield limits for distinguishing the actual width of lines and dots. A caution on spot size constraints based on fidelity of line width <"20 microns is given (other constraints are mentioned). Series expansions within the mathematical definitions reveal that the detection thresholds are simply related to polynomial power integrals (i.e. moments) of both the object profile and the visual system spatial filter function. An argument is presented that a halftone screen can not be free of screen-to-raster artifacts at addressabilities <'-1200 d.p.i. if the screen is not rational tangent. Published limits on the visibility of edge raggedness suggest that stairstepping at addressabilities <-3700 d.p.i. should never be a concern.

The development of an electrostatic printing system that is intended to emulate conventional printing must not only reproduce the color of the ink and the optical density of the printed solid area it must also obtain midtone dot gains and dot gain curves essentially identical to those obtained with conventional printing. In electrostatic printing halftone dot size may be affected by image charge decay which is accelerated in small features, enhanced toner development along edges due to electrostatic fringe fields, and by the transfer of the image to paper. This report will show how these phenomena are dependent on the size of the halftone dots and how they combine to impact dot gain in a liquid electrostatic printing system. Specifically, it is demonstrated that differences observed in midtone dot gains on prints obtained with two different electrostatic masters are primarily the result of toner transfer from the masters to paper, while differences observed in the shape of the dot gain curves on prints obtained from the two masters are primarily due to the charge decay and toner development characteristics for halftone dots of the two masters.

High quality and high work efficiency Direct Digital Color Proofing (DDCP) system is an important demand in pre-press industry. This system requires the peiformnce such as halftone dot reproducibiIity, high-speed output and easy i mage processing.

The photopolymers discussed here utilize a photo-initiated crosslinking mechanism to detackify the imaged regions. The latent image can be rendered visible by applying micron-sized pigmented toner particles to the tacky non-imaged areas of the photopolymer. Because the toner contains a solid plasticizer, it serves as a latent ink particle; heating activates the delayed tack state, characterized by prolonged adhesiveness. The creation of a liquid ink enables the rendered image to be printed onto a paper substrate. Photopolymer components, migrating into the melted toner layer, preferentially adsorb to the polymer/air interface after printing to paper, thus influencing the delayed tack adhesive state, during which excess plasticizer crystallizes. We examine competitive adsorption phenomena at the polymer solution/air interface, via a model four-component system, containing two oligomeric surfactant molecules from the film, both characterized by ethylene-oxide linkages, but distinguished by the fact that one, is linear, while2 the second, is composed of three branches. The concentrations of these long-chain constituents are varied in an otherwise constant bulk polymer solution of triphenylphosphate and the toner polymer simulating typical compositions in the actual melted toner layer. We apply two techniques of surface analysis, SIMS and ESCA (XPS), in order to reveal the identity of the adsorbing film species. Although both molecules are capable of a surface excess, the structure imposed upon the interfacial region is clearly different, depending on the excess surfactant at the surface. This difference is explained by assuming that the linear molecule lies flat on the surface while the branched oligomer adsorbs vertically. Specific SIMS/ESCA signals exhibit spectral intensities that are nonilnearly proportional to the bulk oligomer concentration; from the shapes of the adsorption isotherms, we deduce that the single chain surfactant is displaced at the surface by the branched oligomer.

The Color Rendering project at IBM Almaden Research Center is examining the problem of how to display and print quality color images. For this work we need a printer calibrated to a device independent color space. In this paper, we describe a system for printing CIELAB images on a CMYK printer, and focus particularly on calibration methods. We use tri-linear interpolation to convert CIELAB colors to CMY or CMYK colorants. We obtain the interpolation table by inverting a tetrahedral linear interpolation of a calibration table constructed by measuring printed color patches. Since tetrahedral interpolation has a simple analytical inverse, we can produce the inverted table much more quickly than with the numerical methods needed to invert a multilinear interpolation, even though we have to measure more patches to obtain the same accuracy. To cover the full printer gamut, we found it necessary to add some out-of-gamut entries to the inverted interpolation table. These entries must be obtained by extrapolation, and increase the errors interpolating colors on and near the gamut surface. To date we have calibrated a DuPont 4Cast to print CIELAB colors using CMY colorants. We discuss our results with this calibration, how we propose to add black, and how we fit gamut mapping into the processing.

The uniformity of a color model (space) is a desirable property for various applications, one of which is color imaging and visualization of quantitative information. In such applications, another desirable property is the availability of efficient algorithmic transformations between the given model and other color models (e.g., RGB for CRT displays, XYZ for standard specifications). The uniformity of the CIELUV and CIELAB spaces is limited to local neighborhoods and rather inadequate globally. The transformations from CIELUV and CIELAB to other spaces are mathematically complex and computationally inefficient. The Munsell Book of Color presents true uniformity but covers only a subset of all visible colors ( roughly 1600 uniformly spaced samples). Transformations to other spaces are also complex and require additional interpolations for points not in the original subset. GLHS—the generalized lightness, hue, and saturation color model [7] is a family of LHS models whose members are selected parametrically. A single pair of algorithms provides efficient transformations between all family members and the RGB model. We discuss an optimization process to find a GLHS family member that is the closest approximation to the Munsell Book of Color. We compare the resulting model to the hexcone model, the Munsell Book of Color, and to the CIELUV space.

A method of color correction which matches color printouts to the color monitors used to view the image and the criteria used to develop the method are presented. This method will be contrasted with alternative matching methods, such as tristimulus match. The tristimulus match maps in-gamut colors to the closest color available on the output device. If the color is out-of-gamut, then an algorithm (such as maintain luminance, hue, or saturation) is used to choose a color. When an image changes color in a smooth manner, the in-gamut colors match very closely, and the out-of-gamut colors are pushed to some compromise color. The transition from in-gamut to out-of-gamut color is readily apparent. If the in-gamut colors are adjusted to a less optimal match, the out-of-gamut colors can then be given a better match, providing a smooth transition between in- and out-of-gamut colors. The method presented gives a more pleasing color match than a tristimulus match, if the image to be matched includes colors which are both in-gamut and out-of-gamut for the output device, and is especially useful if the output device has a small color gamut.

Accurate scanning of a colour image, which is absolutely essential for good colour reproduction, can ensure that all relevant information about the colour stimulus of a signal is obtained. This paper introduces a measure of goodness of a set of colour scanning filters. The q-factor of a single colour scanning filter is shown to be a particular case of the measure. Experimental results are presented to justify the appropriateness of the measure.

I consider the problem of numerically computing tristimulus values for a given spectral power density. In particular, I examine the use of interpolatory quadrature rules for the solution of this problem. A good deal of effort has gone into creating tables of weights and abcissas for solving this problem [3]. Wallis [2] has proposed a more sophisticated approach using Gauss quadrature rules. I show that the performance of these techniques can be improved in a well-defined sense, and derive a method based on a new class of quadrature rules. These rules give optimal performance in the sense that they maximize the overall degree of precision while simultaneously minimizing the number of function evaluations.

This paper describes a machine learning algorithm for analyzing the multispectral images of natural scenes. A mathematical training algorithm has been developed to guide the operation of a statistical pattern recognition technique for detecting and extracting the image clusters in a multidimensional feature space. For this purpose, the peak modality of l-D image histograms is selected as the mathematical training criterion. The algorithm is applied to the clusters of the color images of natural scenes in 3—D feature space. During the training process, image clusters are detected in some well—defined decision elements using constant lightness and chromaticity loci of the uniform color space. This gives non—parametric estimates of the clusters' distributions without imposing any constraints in their forms. The linear discriminant method is then used to project simultaneously the detected clusters onto a line for region isolation. This permits utilization of all the spectral properties for object recognition and inherently recognizes their respective cross correlation.

The use of video cameras based on charge-coupled devices (CCD's) for color measurement is a new and growing field which does not fit into any established category of image processing. Just as silicon photodiodes replaced photomultiplier tubes in colorimeters twenty years ago, the CCD array is beginning to replace the silicon diode today. A good colorimeter has a measurement repeatability of the order of 1 part in 10,000, or 0.01 in XYZ or L*a*b*. By contrast, most video frame grabbers are limited to a resolution of 1 part in 256. However, because a color camera can produce in excess of 23 million 8-bit numbers per second, appropriate image processing may be used to improve repeatability and reproducibility dramatically. In this paper, the CCD camera is evaluated from the standpoint of conventional colorimetry. The following points are addressed: requirements of a camera to be used for color measurement; possible physical configurations; sources of noise and repeatability of color measurements; image processing algorithms for colorimetry; RGB-to-XYZ transformation; and requirements for an image processor to be used with a camera-based colorimeter.

Two new techniques are introduced for measuring the color content of printed graphic images with applications to web inspection such as color flaws and measurement of color quality. The techniques involve the development of algorithms for combining the information obtained from commercially available CCD color cameras and colorimeters to produce a colorimeter system with pixel resolution.

Since their earliest market presence in the mid- 19705, xerographic printers have been a valuable asset to the information marketplace. The year 1991 was no exception. Many improvements, enhancements and innovations in xerographic printers were presented by a host of vendors.

Application of the Partially Coherent Imagery method to the Laser Raster Output Scanning (ROS) system with Acousto-Opto Modulator (AOM), appears to lead to a nonlinear Modulation Transfer Function (MTF). When using AOM for Laser ROS, it is necessary to consider not only conventional parameters but also two more conditions, the illumination geometry of AOM and the dynamic image of the sound wave in the AOM. The illumination geometry of the ROS is similar to that of a microscope system. The calculation of partially coherent image is useful for such systems. The most important factor of that calculation is Relative Numerical Aperture (RNA), the ratio between illuminative NA and objective NA. Replacing the laser beam diameter with the illuminative NA, leads to the MTF curve for the static optical system. This system shows a nonlinear response, due to a nonlinearity between light amplitude and intensity. The response for the dynamic image of the sound wave at the AOM, enables to calculate the MTF of the dynamic optical system which has constant velocity error. Combining these and other MTF responses, the total system MTF curves are investigated. The nonlinearity manifests itself by the peak value which is enhanced in comparison to the continuous lighting level. As a total, the response of this system is higher than the response of a conventional ROS system in practical spatial frequency domain.

A scanning optical system employing plastic lenses was studied in order to realize a high resolution laser scanning optical system with simple construction. A simulation program was developed to analyze the effect of lens surface error which was a problem with plastic lenses. As a result we confirmed suitable lens disposition which would make possible a high resolution optical scanning system when employing current high precision plastic lens production engineering. Including this result, compensation of focus shift and correction of aberration were studied. Consequently, the optical scanning system with simple construction was designed, and a small and uniform scanning spot was achieved in our prototype of the optical system.

In a high quality LED printing system, the non-uniformity of the printhead system not only has to be corrected but maintained over an extended period. To further understand the impact of LED printhead aging on realistic high quality Electrophotographic printing system, we have designed and built a LED printhead aging system and used a 400 dpi printhead to simulate actual printing in a control environment (temperature and drive current). We have pushed this study to an equivalent 40 million pages in printing. A representative image suite for document printing is scanned in and used as the actual printing images. The printhead non-uniformity has been previously corrected before the aging experiment starts. During the aging experiment, the printhead pixel irradiance on the film plane was periodically rescanned radiometrically and the exposure non-uniformity of the printhead using the original correction was calculated and compared with a halftone visual detectability curve. Printhead aging due to realistic image printing usage does not appear to create noticeably observable banding problem even after 40 million pages of simulated printing.

This paper discusses an analytical investigation of the motion quality requirement of an image carrier with respect to the given allowable image misregistration between the primary colors. The system analyzed is concerned with a one pass configuration in which the image carrier traverses by sequentially multiple imaging stations in a single pass. Examples of such a system are full width ink jet color printers and thermal transfer color printers. The motion quality of the image carrier, which is the printed substrate in the above examples, impacts directly the positional registration between the colors and in turn the print quality. To characterize the image carrier motion, its traversing velocity is represented by the Fourier sum of the dc and ac components. The amplitude and frequency of the ac velocity component are used as the characteristic measure in the context of motion quality. The analytical expression for motion quality is derived by applying one ac velocity component at a time. The frequency domain spectrum is then generated by sweeping the frequency through the range of interest. An intriguing phenomenon called "error synchroneity" can be shown when the spatial period of the ac velocity component is an integer fraction of the distance between the imagers. At these frequencies, the positional errors of the different color images are offset, regardless of their amplitude. Consequently, the print of the composite images is perceived to be in registration, without color shifts and other associated problems.

Image geometric accuracy in large format color separation quality writing engines or output scanners requires careful control of the opto—mechanical alignment of engine subsystems. Internal drum scanners are particularly sensitive to misalignment of the spinner, drum and spinner linear transport axes relative to each other. Misalignment of these axes leads to significant errors in cross—scan and in-scan pixel placement in the written image. A method is presented for determining alignment sensitivity of internal drum output scanners by tracing only the principal ray through the system and systematically manipulating engine component axes with the ray trace model. The model traces deviated rays through the misaligned system providing data on placement of pixels in the written image. This method provides an efficient means of determining system sensitivity to various alignment conditions and allows system tolerances to be apportioned appropriately.

Printers process large quantity of data when printing. For example, printing on an A3 size (297 mm 420 mm) at 300 dpi resolution requires 17.4 million pixels, and about 66 Mbytes in a 32-bits/pixel-color image composed of Yellow (Y), Magenta (M), Cyan (C) and Black components. Containing such a large capacity of Random Access Memories (RAMs) in a printer causes an increase in both cost and the size of memory circuits. Thus, image compression techniques are examined in this study to cope with these problems. A still-image coding, being standardized by JPEG (Joint Photographic Experts Group)', will presumably be utilized for image communications or image data bases. The JPEG scheme can compress natural images efficiently; it is unsuitable for text or Computer Graphics (CG) images for degradation of restored images. This scheme, therefore, cannot be implemented for printers which require good image quality. We studied codings which are more suitable for printers than the JPEG scheme2. Two criteria were considered to select a coding scheme suitable for printers: i.e., (1) no visible degradation of input printer images; (2) capability of image edition. Especially in terms of criteria (2), a fixed-length coding was adopted; an arbitrary pixel data code can be easily read out of an image memory. Then we implemented an image coding scheme in our new sublimation full color printer. Input image data are compressed by coding before being written into an image memory.

In black-and-white printing the page image can be represented within a computer as an array of binary values indicating whether or not pixels should be inked. The Boolean operators ofAND, OR, and EXCLUSIVE-OR are often used when adding new objects to the image array. For color printing the page may be represented as an array of "continuous tone" color values, and the generalization of these logic functions to gray-scale or full-color images is, in general, not defined or understood. When incrementally composing a page image new colors can replace old in an image buffer, or new colors and old can be combined according to some mixing function to form a composite color which is stored. This paper examines the properties of the Boolean operations and suggests fullcolor mixing functions which preserve the desired properties. These functions can be used to combine colored images, giving various transparency effects. The relationships between the mixing functions and physical models of color mixing are also discussed.

This paper presents a methodological approach for integrating non-colorimetric scanners with CIE standards as a means toward a device independent process. The calibration is aimed at reflected samples by employing photographic, thermal transfer, and xerographic prints. Correlations between scanner responses to CIE standards are established through a common test object using a two-step process of the gray balancing and the matrix transformation. A series of polynomials, ranging from a three-term linear combination to a twenty-term cubic equation, is used for converting device values to a CIE color space. The ability to fit colors that are not in the training set by a polynomial is examined. Results indicate that lower order polynomials fit colors equally well whether a color is in the training set or not, but the accuracy of interpolation decreases as the number of terms in the polynomial increases. We study the generality of this calibration method with respect to input materials. The transformation is material-dependent. Within the experimental uncertainty, however, there exists a unified transfer matrix for photographic materials and another one for paper substrates. Finally, we extended this method to deal with the mismatched illuminants for viewing and calibration. An empirical white point conversion method is proposed and tested; good approximations to the measured results are obtained when the interchange of illuminants occurs.

In response to the recent market demand of faithful conversion of CRT iniages into the form of hardcopies, Fujix Pictrography 2500, which is the latest version of Fuji Photo's digital printers, is equipped with a new type of image processing unit. This unit is characterized by the capability of minimizing the effect of fluctuating monitor characteristics as well as that of non-linearly transforming colors based on a new color mapping algorithm which takes into account chromatic adaptation of human vision. Due to this image processing unit, differences in the color appearance between CRT and hardcopy images have markedly reduced compared to those with the current Pictrography 2000.

1.1 Premises It is an advantage of open publishing systems that it is possible for the user to put together parts of different suppliers to a complete system. Standardized interfaces provide the links between the different parts of a system. One of the key technologies for output is PostScript, the well-known page description language, developed by Adobe systems. It has become a de-facto standard for computer publishing applications. The main advantage of PostScript is its device independence: Any document can be printed on any PostScript output device in any available resolution. The idea of device independence, however, leads to some problems: If there are problems concerning output quality, it is very difficult to detect the source of these problems. Without additional help it is often impossible to find out which part of the system (scanner, software, output engine, the developing process) causes the problem. 1.2 Objectives Until now, Computer Publishing systems lacked control systems like those used for offset printing. The objectives of the VFG-GLV-PostScript-targets were to develop test targets, which enable any user to evaluate and maintain the quality of any output device (printer or imagesetter). These test targets should be supplier independent. 1.3 Developing the test targets Developing the test targets started early in 1990 at the institute for Graphic arts in Vienna (Graphische Lehr- und Versuchsanstalt, GLV). The first protoypes were used at CeBIT 1990 for an imagesetter comparison. In 199 1 the concept was refined, giving two test targets: The VFG-PTF and the VFG-PQF. After field testing in 199 1 they are available around the world since September1991. The test targets were developed using direct PostScript programming. Main parameters of the output device are interrogated and are used to build test targets suited for a specific device resolution on the fly. Some test targets are device dependent and show output quality for a specific device at a specific resolution. This is the advantage of a test target, created by a program. It allows to deal with specific devices. The test targets ar therefore suited to test imagesetters at high resolutions as well as low-res laserprinters. 1.4 VFG-PTF: Quality evaluation The VFG-PTF test target helps evaluate the quality of a PostScript printer or imagesetter. Its size of 20x20 cm allows for a large quantity oftest targets. The test target checks resolution, screen angles, density, and the quality of mechanical and optical system. The evalutation of the test targets will be done visually or using a densitometer. 1.5 VFG-PQF: Quality maintenance The VFG-PQF test target is used for job quality control. It is pasted into a PageMaker-, Ventura Publisher- or XPress-job and will be sent to the printer with the job. This allows quality control for every page processed

The ink-jet marking process offers several unique opportunities for producing quality hard-copy images. There are, however, certain limitations and requirements of the technology that must be taken into account when developing imageprocessing procedures and algorithms for ink-jet printing systems. This paper describes a number of issues that set ink-jet apart from many of the other marking processes. For example, ink-jet can be treated as a truly "binary" marking process. Thus, single isolated pixels are easily and reproducibly formed on the marking substrate. Halftoning procedures have been developed that take advantage of this attribute to produce more gray levels for a given resolution. Ink coverage on paper, however, must often be limited to < 200% . Also, the perceived color will be dependent on the order in which the colors are delivered to the marking substrate. Examples illustrating these and other concerns are given. Optimal image-processing procedures for the ink-jet marking process can be developed based on an understanding of these and other ink-jet specific issues.

The FUJI XEROX Able 1301a is the first single pass highlight color copier, that produces two color copies through a single Xerographic process. The Able 1301a provides superior copy quality at a low cost through the following new unique technologies. 1. Color flag method: Color flag method reduces image data size. Instead of using a l6bit wide data path, the internal image data consists of a 1-bit flag, that represents color (black or color) and 8-bit luminance data. 2. Pulse width modulation (PWM) technology in delay line method: Delay line type PWM does not requires a high frequency video clock which are used for the conventional digital shift clock type PWM. Besides, the circuits for the delay line PWM cost less and provides less drift than those for analog circuit PWM. 3. Line screen forming method: The technique used to generate the screens enables selectively printing with 200 or 400 line screens per inch for Photo and Text modes using a simple circuit. 4. Statistical process for background suppression: This technique enables background detection and background area calculation to effectively produce copy without high background with various types of originals.

The system of C tristimulus coordinates is an accepted standard for the device independent exchange of color information. It is possible to print colors that are given in this system with low cost bilevel color printers of different print technologies. Results achievable with the Mannesmann Tally MT7400 C thermal transfer and the Mannesmann Tally MT92 C bubble jet printers are presented.

Electronic reprographic systems are those which reproduce images through digital means. These systems offer the possibility of more flexible tonal control over pictorial reproduction than earlier analog systems. One challenge of electronic reprographics is to design a method of pictorial tone reproduction control that is simple to operate and that does not require operator knowledge of image processing. Described here is one such method which has been applied to Xerox's DocuTech Production Publisher. All electronic reprographic systems have three major components: a digital scanner, an image processor, and a digital printer. The imaging characteristics of all of these components could be altered to effect the system's tone reproduction. With the described method, scanner and printer imaging characteristics are held rigidly constant through rigorous process controls. This permits direct calculation of image processing characteristics to meet the tone reproduction requirements specified by the customer. The method allows lower cost hardware and faster processing by a reduction in the number of gray levels per pixel. A userfriendly interface and the availability of multiple halftone screens also contribute to meeting customer requirements.